Café Opening Hours: Which Ones Actually Pay for Themselves?

Most cafés pick their hours once, in the week before they open, and never revisit them. Seven to eleven sounded right, or it matched the unit next door, or the lease said so. Two years later the café is still unlocking at seven for a handful of customers and still standing behind a lit counter at half past ten because that is what the sign says. Nobody decided to keep doing it. It just never came up as a question. It is worth making it one, because the first and last hours of a trading day are where a café is most likely to be paying more to be open than it takes.

How do you tell which hours are earning?

Pull your sales by hour of day for the last four to six weeks — not by day, by hour. Then split it by weekday, because a Tuesday and a Saturday are different businesses trading under the same name, and averaging them hides both. What you want in front of you is a simple grid: each hour of the day, each day of the week, and what came through the till in it.

Read two numbers per hour, not one. Takings tell you what arrived. Transactions tell you whether anything was happening — six customers spending AED 300 between them is a very different hour from thirty customers spending the same, and only the transaction count separates them. An hour with money but almost no transactions is usually one big order that could have been taken at any time. An hour with transactions but little money is often the one your team dreads and your accountant likes.

Four weeks is the minimum that stops one unusual day from deciding the answer. If your café has a strong seasonal shape — and in the UAE most do — check the same hours against the equivalent weeks last year before you commit, because the year moves in four distinct periods and the hour that dies in August often comes back in November.

What does an open hour actually cost?

This is the part owners skip, and it is why marginal hours survive so long. The instinct is to compare the hour's takings against zero: anything is better than nothing. But the hour is not free, and the comparison that matters is takings against what it costs to have the door open.

Add the first three and you have a rough floor: the takings an hour needs to clear before it is worth unlocking for. Most cafés find one or two hours a week that sit under it consistently, and they are almost never the ones the owner expected. If you do not keep a waste log yet, the two-minute version is enough to make the last-hour line real rather than guessed.

Should you close early or open later?

They are not the same decision, and the answer is usually different at each end of the day. The morning hour tends to be worth more than it looks. Early customers are disproportionately regulars, they buy on their way somewhere rather than on a whim, and they are the hardest to win back once they have found somewhere else that is open when they pass. Losing the seven o'clock customer often costs you their eight o'clock visit too.

The closing hour behaves differently. Late trade is usually made of people already sitting down, which means the hour frequently shows takings that were actually earned earlier — the bill lands at ten for drinks poured at nine. Strip out anything that was ordered before the hour started and the last hour on its own is often much thinner than the report first suggests.

So test them differently:

• At the close, try trimming thirty minutes rather than an hour, on the two weakest weekdays only, for a month. Small enough that regulars barely notice, long enough that the numbers mean something.
• At the open, resist trimming first. Try moving the work instead — one person opening rather than two, prep pulled forward, the second barista arriving for the rush rather than for the sunrise.
• Either way, change one thing at a time. Cutting the morning and the evening in the same month leaves you unable to say which one the drop came from.
• Watch the following hour, not just the hour you changed. The real cost of closing at nine instead of ten shows up in whether the eight o'clock hour also shrinks.

What if the quiet hour is a staffing problem, not a demand problem?

Before you shorten the day, check whether the hour is genuinely dead or simply badly covered. An hour with one exhausted person on the floor, no prep done and a queue that forms every time somebody orders food is not a measurement of demand — it is a measurement of the rota. The honest test is whether the hour was ever properly staffed and still failed to pay. If it wasn't, fix the rota first and re-read the numbers a month later; staff scheduling for a small café works through building the rota around the peaks rather than across the day.

The same applies in reverse at the busy end. If your peak hour is losing customers to a queue, extending your closing time adds nothing — the money is walking away at half past eight, not at half past ten. Finding the bottleneck in your rush is a far better use of the same effort, and it usually costs nothing to fix.

The constraints you cannot schedule around

Your hours are not entirely yours to set. A mall or a hotel lease commonly fixes trading hours, and the permitted range can differ from the mall's own headline opening times — read the clause rather than assuming, and ask the centre management in writing if it is ambiguous. Your team's contracted hours matter too: shortening the day does not automatically shorten anyone's contract, and the daily and weekly limits in the federal labour law still apply to whatever shape you land on. If your unit sits in a residential building, noise and delivery timing may be constrained separately. None of this stops you optimising — it just means the change goes past the lease and the contracts before it goes on the door.

Where MidaOne fits

The whole exercise depends on having hour-by-hour sales you can trust, and that is a till job rather than a spreadsheet job. MidaOne records every sale as it happens, so sales by hour, by day and by item are already sitting there when you sit down to look at the question, alongside a busiest-hours chart on the accounting page. Shift closes show what was expected against what was counted, which is what tells you whether a thin hour was thin or simply not rung up. It runs on the phone or tablet you already have, on one flat price of AED 200 a month with unlimited devices.

Frequently asked questions

The reason marginal hours survive is that nobody is ever asked to defend them. They were set before the café had a single customer, and they have been carried forward on the assumption that being open is always better than being closed. It usually is — but not always, and you already own the data that says which. An hour with your own hourly report is enough to find out, and the answer is often thirty minutes at one end of the day that were costing you a barista's wage and buying almost nothing back.